sharpen.md

Sharpen

Sharpens an image through the unsharp masking method. Applies a gaussian blur which is subtracted from an exaggerated version of the starting image.

plantcv.sharpen(img, ksize, amount=1, threshold=0, sigma_x=0, sigma_y=None, roi=None)

returns sharpened image

• Parameters:
  • img - RGB or grayscale image data
  • ksize - Int, Tuple, or numpy.ndarray. Used to make a tuple of kernel dimensions, e.g. (5, 5), which must be odd integers.
  • amount - Integer describing amount of sharpening, higher numbers will sharpen more.
  • threshold - Integer cutoff on low contrast, contrasts lower than this will be removed.
  • sigma_x - standard deviation in X direction; if 0 (default), calculated from kernel size
  • sigma_y - standard deviation in Y direction; if sigma_Y is None (default), sigma_Y is taken to equal sigma_X
  • roi - Optional rectangular ROI as returned by pcv.roi.rectangle within which to apply this function. (default = None, which uses the entire image)
• Context:
  • Used to reduce blur in an image

Original image

Sharpening Image

# Apply sharpening within an ROI to show differences
roi = pcv.roi.rectangle(img, 200, 0, 335, 200)
sharp1 = pcv.sharpen(img, (5, 5), amount=1, roi=roi)

# Higher amount of sharpening will look more dramatic
sharp5 = pcv.sharpen(img, (5, 5), amount = 5, roi=roi)

Sharpen (ksize = (5,5), amount=1, roi=roi)

Sharpen (ksize = (5,5), amount=5, roi=roi)

Source Code: Here